Skip to main content

Full text of "Bergman Kernel from Path Integral"

See other formats


Preprint typeset in JHEP style - HYPER VERSION 



RUNHETC-2008-15 
ITEP-TH-37/08 



Bergman Kernel from Path Integral 



Michael R. Douglas^'^ '^ and Semyon Klevtsov^'^'^ 

^ Simons Center for Geometry and Physics, Stony Brook University, 
Stony Brook, NY 11794-3840, USA 

^ NHETC and Department of Physics and Astronomy, Rutgers University, 
Piscataway, NJ 08855-0849, USA 

^ I.H.E.S., Le Bois-Marie, Bures-sur-Yvette, 91440, France 
4 ITEP, Moscow, 117259, Russia 

mrdSphysics . rutgers . edu, klevtsovOphysics . rutgers . edu 

Abstract: We rederive the expansion of the Bergman kernel on Kahler manifolds devel- 
oped by Tian, Yau, Zelditch, Lu and Catlin, using path integral and perturbation theory, 
and generalize it to supersymmetric quantum mechanics. 

One physics interpretation of this result is as an expansion of the projector of wave functions 
on the lowest Landau level, in the special case that the magnetic field is proportional to the 
Kahler form. This is relevant for the quantum Hall effect in curved space, and for its higher 
dimensional generalizations. Other applications include the theory of coherent states, the 
study of balanced metrics, noncommutative field theory, and a conjecture on metrics in 



black hole backgrounds discussed in |24|. We give a short overview of these various topics. 
From a conceptual point of view, this expansion is noteworthy as it is a geometric expansion, 
somewhat similar to the DeWitt-Seeley-Gilkey et al short time expansion for the heat 
kernel, but in this case describing the long time limit, without depending on super symmetry. 



Contents 



|l]. Introduction ^ 

|2|. Background ^ 

|2.1| Particle in a magnetic field ^ 

The lowest Landau level ^ 

Applications in Kahler geometry H 



|3|. Non-supersymmetric Bergman Kernel 10 
|3.1| Density matrix 10 
|3.2| Weyl-ordering counterterm 11 



3.3| Normal coordinates, free action and propagators 11 
3.4 Perturbation theory. First Order 13 
Perturbation theory. Second Order 14 



^ Af = (1,1) Super symmetry 16 

[4.1| Action, symmetries and propagators 16 



4.2 Perturbation theory 17 



H]. N = {2, 2) Supersymmetry [18 
|5.1| Action, symmetries and propagators 18 



5.2 Perturbation theory 19 

H. Conclus ions |19| 

1^ Curvatures [20 

g. Hamilt onian 1211 

in. Integrals [22 



1. Introduction 

A prototypical topic at the interface of geometry and theoretical physics is the study of 
quantum mechanics in curved space, i.e. on a Riemannian manifold M [0, ^, ^, Many 
results in this area are of great interest both to physicists and to mathematicians, with 
some examples being the DeWitt-Seeley-Gilkey short time expansion of the heat kernel, 
and the relation between supersymmetric quantum mechanics and the Atiyah-Singer index 
theorem ^, 0- 



- 1 - 



A more recent result which, although not well known by physicists, we feel also belongs 
in this category, is the expansion for the Bergman kernel on a Kahler manifold developed 
by Tian, Yau, Zelditch, Lu and Catlin ^, 11]. It applies to Kahler quantization 



and gives an asymptotic expansion around the semiclassical limit. This has many uses in 
mathematics 12, |l^, 14 1; see the recent book |15]. 



Here we will provide a physics derivation of the asymptotic expansion of the Bergman 
kernel using path integrals, and explain various possible applications of this result. 

In physics terms, perhaps the simplest way to define the Bergman kernel is in the 
context of quantum mechanics of a particle in a magnetic field, in which it is the projector 
on the lowest Landau level. It is not hard to see that the limit of large magnetic field is 
semiclassical, so that one can get an expansion in the inverse magnetic field strength using 
standard perturbative methods. 

Our basic result is to rederive the Tian-Yau-Zelditch et al expansion as the large time 
limit of the perturbative expansion for the quantum mechanical path integral. We also 
generalize it to = 1 and N = 2 supersymmetric quantum mechanics. 

Let us state the basic result for (nonsupersymmetric) quantum mechanics. We consider 
a compact Kahler manifold M, and a particle in magnetic field, with the field strength 
proportional to the Kahler form on the manifold 

Fij ~ LOij. (1.1) 

One can show (see below) that, just as for a constant magnetic field in flat space, in this 
situation the spectrum is highly degenerate, splitting into "Landau levels." Let the lowest 
Landau level (LLL or ground state) be A^-fold degenerate with a basis of orthonormal wave 
functions 'ipi{x), then we define the projector on the LLL as 

N 



1=1 

We could also regard this as a density matrix describing a mixed state in which each ground 
state appears with equal weight, describing the the zero temperature state of maximum 
entropy. 

We then consider scaling up the magnetic field by a parameter k, as F ^ kF. Note 
that on a compact manifold, F must satisfy a Dirac quantization condition; thus we take 
A; = 1, 2, 3, . . . In the large k limit, the diagonal term then satisfies 

p{x,x) ~ fe" (^1 + Ai? + g QAii + A|Riemp - i|Ric|2 + ii?^^ + Oi{h/kf)^ (1.3) 



as an asymptotic expansion (see Appendix A for the precise definitions of different 
terms here). 

In some ways this expansion is similar to the well known short time expansion of the 
heat kernel, but note that it is a long time expansion, because it projects on the ground 
states. Unlike other analogous results for ground states, it does not require supersymmetry, 
either for its definition or computation. Of course, similar results can be obtained for 



- 2 - 



supersymmetric theories, our point is that that they do not depend on supersymmetry 
(whether they depend ultimately on holomorphy is an interesting question discussed below). 

There are various other physics interpretations of this result. One familiar variation 
is to regard M as a phase space, and try to quantize it, following Berezin [16|. As a 



phase space, M must have a structure which can be used to define Poisson brackets; it 
is familiar |17] that this is a symplectic structure, i.e. a nondegenerate closed two-form 



oj. The definitions we just gave are then the standard recipe of geometric quantization 
P^ . They lead to a finite dimensional Hilbert space, whose dimension is roughly the phase 
space volume of M in units of {2'kK)'^ . In this interpretation, the parameter k plays the 
role of and thus the large k limit is semiclassical. 

From this point of view, it is intuitive that one should be able to localize a wave 
function in a region of volume (27r^)" ~ 1/^", and thus in the large k limit the density 
matrix p{x, x) should be computable in terms of the local geometry and magnetic field near 

X. 

To do this, given a point z G M, one might seek a wave function ipziz') which is peaked 
around z. Given an orthonormal basis for 7i, a natural candidate is 



This is a coherent state (in the sense of |ll8| ). It can be used to define the symbol of an 
operator, the star product [^], and related constructions. In this context the Bergman 
kernel is the "reproducing kernel" studied in pQ], see [jl^ for a review. For recent work 



on applications of the Bergman kernel to quantization of Kahler manifolds see |21, 22 1. 
Another recent paper discussing the topic is ||2^ . 

Our original interest in this type of result came from the study of balanced metrics 
in and a conjecture about their relevance for black holes in string theory stated in 



1 24]. However, after realizing that these results and techniques do not seem to have direct 
analogs in the physics literature and could have other applications, we decided to provide 
a more general introduction as well. 

2. Background 

Let us give a few mathematical and physical origins and applications of this type of result. 
2.1 Particle in a magnetic field 

We consider a particle of mass m (which later we set to one) and charge k, moving on a 
2?7--dimensional manifold M which carries a general metric gij , and a magnetic field Fij . It 
is described by a wave function ip{x;t) which satisfies the Schrodinger equation, 

^ ^^Di ^g'^ Dji^ = EiP, (2.1) 

where Di = idi + kAi is the covariant derivative appropriate for a scalar wavefunction with 
charge k, and E = ihd/dt. If M is topologically nontrivial, as usual we need to work in 



-3- 



coordinate patches related by gauge and coordinate transformations, and this expression 
apphes in each coordinate patch. We can of course also consider the time-independent 
Schrodinger equation with E fixed, and seek the energy eigenstates H'il>i{x) = Eiipi. 

Let us now consider the limit of large magnetic field or equivalently large k. The case 
of two-dimensional Euclidean space gij = 6ij with a constant magnetic field Fij = Beij is 
very familiar. The energy eigenstates break up into Landau levels, such that all states in 
the rth level have energy Ei = hkB{l + \)/fn. Within a Landau level, one can roughly 
localize a state to a region of volume h/kB. 

These results can be easily generalized to d = 2n dimensions. Choose coordinates such 
that the magnetic field lies in the 12, 34 planes and so forth, and B12 > 0, i?34 > etc. 
Then, considering the lowest Landau level (LLL) we have 

E=^{Bu + -.. + B2n-l,2n), (2.2) 

Zni 

with states localized as before within each two-plane. 

In a general metric and magnetic field, while one might not at first expect this high 
degree of degeneracy, it is still possible. When the magnetic field is much larger than the 
curvature of the metric, the intuition that wave functions localize should still be valid. 
Then, we might estimate the energy of a wave function in the lowest Landau level localized 
around a point x as Eq. ( |2.2| ), where the components -B12, -B34 and so on are evaluated in 
a local orthonormal frame. If the energy E in Eq. ( p.2| ) is constant, then all states in the 
LLL will be degenerate, at least in the limit of large k. 

The proper generalization of the splitting of the components of B Eq. ( |2.2| ) for noncon- 
stant magnetic fieds seems to be, that the field strength should be nonzero only for mixed 
components Faa, with Fab = Fab = in the complex coordinates z", (a, a = 1, ...,n) on 
the manifold. Mathematically it means, that the underlying line bundle is holomorphic. 
In this case, the argument can be sharpened by using the identity 

[D,,D,]=Fij. 

to rewrite the Hamiltonian as 

H = g''~^Fa-a+9''~^DaDa. 

This makes it clear that if the following combination is constant 

g'^^'Faa = const (2.3) 

every wave function satisfying 

= Dai; (2.4) 

will be degenerate and lie in the LLL. This argument can hold away from the strict k ^ 00 
limit. 

The condition (|2.3| ) is known as hermitean Yang-Mills equation, and is essentially 
equivalent to Maxwell equation in the case F^''^ = 0. Recalling, that the metric coefficients 



- 4 - 



on the Kahler manifold are related to the Kahler form u as Qaa = —i^aa, 9aa = i^aa, one 
can see that our choice of the magnetic field strength 

Faa = kgaa (2.5) 

does satisfy the condition (p. 31) . 

In fact the previous argument relies only upon the Maxwell equations and the condition 
i^2,o _ This suggests that there exist more general magnetic field configurations, than 
( |2.5| ), for which the LLL is still highly degenerate and the expansion in large magnetic 
fields, analogous to ( |1.3D , is possible. For example this includes the case when 6^'^(M) > 1. 
We will elaborate this question in the future publication. 

The previous physical condition for the field strength is equivalent to the mathematical 
condition that M be a complex manifold with complex structure J = B. For a tensor Jj 

to be a complex structure, it must satisfy the conditions = —1 and = \7[iJ^. The 
first is manifest, and given the expression for J in terms of the vector potential 

so is the second. 

Now, a standard trick to simplify the equations Eq. (|2.4D, is to do a "gauge transfor- 
mation" with a complex parameter 9{x). While at first this might seem to violate physical 
requirements such as the unitarity of the Hamiltonian, in fact it is perfectly sensible as long 
as we generalize another ingredient in the standard definitions, namely the inner product 
on wave functions. Explicitly, we define the wave function in terms of another function 
s{x), as 

4,{x) = e^'=^(")s(x), Dai^ix) = e^'^^^") {ida + kAa - kda9)s{x). (2.6) 

This would be a standard U{1) gauge transformation if 9{x) were real. By allowing complex 
0{x), and assuming 

{i.e. F^'"^ = 0), we can find a transformation which trivializes all the antiholomorphic 
derivatives, 

D-a ^ B-a. (2.7) 

In this "gauge," wave functions in the LLL can be expressed locally in terms of holomorphic 
functions. The only price we pay is that the usual inner product, 

J M 

turns into an inner product which depends on an auxiliary real function, 

/i(x) =e-2i'^^(^), (2.8) 



as 



'M 



y/Tj h'^ix) s{x) s'{x) 



- 5 - 



Taking into account the gauge transformations between coordinate patches, the s{x) are 
holomorphic sections of a holomorphic Une bundle L''. 

In mathematics, one would say that s{x) is a section of evaluated in a specific 
frame, while the quantity h^{x) defines a hermitian metric on the line bundle L^. 

2.2 The lowest Landau level 

Since we have made a one-to-one correspondence between LLL wave functions and holo- 
morphic sections of the line bundle, we can now find the total number of LLL states, which 
we denote A'^. The number of holomorphic sections d\niH^{L^) can be determined for large 
k by the index formula ||6|, ^ 

N = d\mH^{L'')= f A Td(M) = aoA:" + aiA;"-^ + . . . (2.9) 

J M 

where the coefficients are given by certain integrals involving the curvature of the metric. 

Now, given that there is a large degeneracy of ground states and thus a nontrivial 
LLL, it becomes interesting to study the projector on the LLL, or in other words the LLL 
density matrix 

i;E,=Eo 

If we shift H to set the ground state energy Eq = 0, it can also be defined as the large 
time limit of propagation in Euclidean time. Regarded as a function on two variables, the 
projector P becomes the Bergman kernel 

Pix.x') = lim (x|e~"^^|x')- 

Thus it can be defined as a path integral by the standard Feynman-Kac formula. 

The standard example in which the projector on LLL appears in physics is the Quan- 
tum Hall Effect, see for a review |2^]. In the simplest case, one studies the dynamics of 
electrons on a two-dimensional plane with a constant orthogonal magnetic field. At low 
temperatures and high values of the field only the lowest energy level is important. It 
is also interesting to consider a partly filled ground state, with number of fermionic par- 
ticles K < N. In this case one has to introduce a potential V, then particles form an 
incompressible droplet, whose edge dynamics is of particular interest. 

In recent years this problem has been much generalized; to Riemann surfaces in |26| 



and references therein, while higher dimensional examples include the case of S |27], W 



!>4 



\2§ and CP" |9|; see also Q for a review. 

The case of CP" is the first nontrivial case in which we can make contact with the 
results of this paper. The choice made in for the U{1) background field is 



Fa-a ~ Ra-a, (2.10) 

proportional to the Ricci tensor. Since for CP" the Ricci tensor is equal to the Kahler 



metric, one immediately recognizes Eq. (2.10) as the physical condition on the magnetic 



-6- 



field Eq. ( |2.5D . Using the local complex coordinates zi^ . . . ,Zn, the LLL wave functions can 
be constructed explicitly as 



^1 ^r. 



up to a normalization constant |3C]. As in Eq. (|2.6D it has the form of the holomorphic 
function, weighted by a metric of the line bundle Eq. (^^), or, equivalently, the magnetic 
potential. 

The dynamics of the droplet is characterized in the following way. One starts with 
diagonal density matrix po with K states occupied, then the fluctuations, preserving the 
number of states are given by unitary transfomation po —>■ p = UpoW , and the equation 
of motion is the quantum Liouville equation 

The form of the droplet is determined by the form of the minima of the confining potential. 



In |2£] the case of spherically symmetric potential V = V{r = zz) was studied. In the limit 
of large number of states (i.e. large magnetic field) and large number of fermions K < N 
the density matrix has the form 

p{r') = e{r'-Rl), 

where Rd is the radius of the droplet and is the step function. In other words the density 
matrix is equal to constant in the region, occupied by the droplet. This is due to the fact, 
that the LLL is only partly filled, otherwise it would have been constant everywhere in 
space. The condition of the constant density matrix (Bergman kernel) turns out to have 
interesting consequences. 

The edge dynamics of the droplet is described by Chern-Simons type action in higher 
dimensions p9| . 

One can generalize the above construction to nonabelian background gauge fields. 
Since CP" = SU{n + l)/U{n) and Lie algebra of ?7(n) = [/(I) x SU{n), then in addition 
to C/(l) gauge field one can also turn on SU{n) gauge field. In ||2^ the case of constant 
SU (n) gauge field was considered, so that the field strength is proportional to the SU (n) 
component of the Riemann curvature. The wave functions ( p. 11 ) as well as the density 



matrix now carry additional indices, corresponding to SU{n) representation. 

The similar generalization of the Bergman kernel was considered recently in In 
addition to the line bundle L one can consider more general hermitian vector bundle E 
with corresponding connection with the curvature R^ . Then the corresponding Bergman 
kernel is given by the large time limit of the exponential of Dirac operator D squared, for 
which the expansion analogous to ( |1.3[ ) exists 



p{x)= lim e-™'(x,x) = ifc^ + A;"-^^^- + + •••• 

T — >oo 2 



The second term was computed in [31|. It would be interesting to make further contact 



between these results and the higher dimensional Quantum Hall Effect. 



-7- 



2.3 Applications in Kahler geometry 

The original mathematical motivation for this development, usually attributed to Tian and 
to Yau, seems to have been to use Bergman metrics to study the problem of approximation 
of Kahler-Einstein metrics, which by definition have Ricci tensor proportional to the metric 
itself, on complex manifolds. 

In Q Tian considered an algebraic manifold M of complex dimension n, embedded in 
some projective space CP^, N > n. Turning on the magnetic field is equivalent to con- 
sidering a bundle L, or it's A;-th power L'^ for magnetic flux proportional to k, whose 
choice corresponds to a choice of "polarization" on M. In local complex coordinates 
z"-, z"", a, a = 1, ... n the Kahler form ujg of the metric gaa is defined as Ug = igaadz"- A dz°-. 
The Kahler metric, polarized with respect to L, has the associated Kahler form ujg in the 
same class as the Chern class Ci{L) of L. A particularly useful choice of LOg is to take it 
to be equal to the curvature of the line bundle (magnetic field strength) . If the hermitian 
metric of L is h[z, z) then for the metric is and it's curvature is 

kga-a = Fa-a = -OaBa log /l^ (2.12) 

exactly as in Eq. (2.5). 



Consider next some orthonormal basis sq{z), . . . , S]\fi_{z) on the space H°{M,L'') of aU 
global holomorphic sections of L'^ 

{Sa,Sp)= ^/gh^SaSp = 5af3- (2.13) 

Jm 

One can think of Sa as of projective coordinates on CP^'' . Therefore a particular choice of 
the basis of sections defines a particular embedding of the manifold M into CP^'^ (different 
choices of the basis are related by PGL{Nk + 1) transformation). The standard metric on 
the projective space is the Fubini-Study metric gps = ddlogY^^ I-Sq^. One immediately 
realizes that ^-multiple of it's pullback \gFs\M to M is in the same Kahler class as the 



original metric g (2.3), since 




^9fs\m = g + ^ddlog 

and the expression inside the logarithm is a globally defined function. This metric is called 
the Bergman metric. In Tian proved, that as A: — > oo, the Bergman metric converges 
to g (at least in topology). This result opened a possibility of approximating the 
Kahler-Einstein metrics by the Bergman metrics. 

It is interesting to look at the structure of the "density of states" function 

Nk 

Pk{z) =h^^^ Sa{z)Sa{z). 
a=0 

Zelditch P| and Catlin |^] showed that there is an asymptotic expansion of the density 
function in 1/k in terms of local invariants of the metric g, such as the Riemann tensor and 



-8- 



its contractions. These invariants were computed by Lu ijlTI up to third nontrivial order 
in \/k with the following result up to the second order in 1/k 



Pf^[z) = k^ + \k''-^R + k''-^ ("iAi^+i-lRieml^-ilRicp + lii^V 0(^n-3) (2.14) 
2 \o 24 D " / 

The computation is based on the global peak section method, developed by Tian which 
is a technique to approximate sections of line bundle for large values of k. Another methods 
to derive this results are the heat kernel approach of [^] reproducing kernel approach of 
1 32]. Let us also mention their importance for the proof of holomorphic Morse inequalities 



|33, 31|. 



In this paper, we reproduce the expansion ( 2.14 ) by taking the large time limit of 
the quantum mechanical path integral for a particle in magnetic field. The function is 
nothing but the diagonal of the density matrix on the lowest Landau level. 

Based on the results of ^, 10, 11 1 Donaldson suggested to study the metrics with 
constant density function 

dim H^(M,L^) 

Pk\z) = const = ■ 



VolM 



Solving the previous equation for and plugging back to the orthonormality condition 



Eq. ( 2.13 ) we get the equation 

dim H°{M,L'') f Sasp 



VolM yM^E,M7"'^"'' 

on the sections of the line bundle. This is the orthonormality condition for the basis in 
H^{M,L^). The embedding M CP^S satisfying this condition, is called "balanced" 



p4I c^iid the corresponding Kahler metric gaa is the "balanced metric" (see [35| for the 



first appearance of this concept). Using the expansion from Eq. ( |2.14| ) Donaldson was 



able to show |12, 13, 14| that under assumption of existence of constant scalar curvature 
metric, the metric, satisfying previous equation, approaches the metric of constant scalar 
curvature as A; ^ cxo. In |lj] an iterative procedure was proposed to construct these metrics 
numerically. One starts with an arbitrary choice of basis, parameterized by a hermitian 
matrix Gap, and defines the following integral operator 

_ dimgO(M,L^) f sg-sp 
The fixed point of this operator T{G) = G corresponds to the balanced embedding. It was 



shown in |12, 14| that for any initial choice of the matrix G, the iterative procedure for T 
converges to the balanced embedding. This construction was recently used for approximat- 
ing the Ricci flat metrics on Calabi-Yau hypersurfaces in projective spaces and finding 
numerical solutions to the hermitian Yang-Mills equation on holomorphic vector bundles 



-9- 



3. Non-supersymmetric Bergman Kernel 



3.1 Density matrix 

The euclidean path integral for a particle on a 2n-dimensional Kahler manifold M with 
the magnetic field is 

Jx{U)=x, t,<t<tf 

Here we assume that Fab = F'ab ~ work in the anti-holomorphic gauge Aa = for the 
gauge connection^. We also set the magnetic field strength to be aligned with the metric 

= daA-, = kg^i = kdad-,K, (3.2) 

as in Eq. (p.5|), with K = — log/i being the Kahler potential for the metric. The physi- 
cal reason for this choice of the magnetic field strength was outlined in the introduction 
(namely to get a highly degenerate ground state). 

Our goal is to compute the value of the density matrix (3.1) on the diagonal 



and in the lowest Landau level, i.e. in the large time T = tj- — tj— s-oo limit. However, 
it turns out that one cannot take this limit before doing the computation. If one does 
this, then the kinetic term in the action is suppressed, and one obtains the result 1 for the 
functional integral, as is easy to check in first order in h. Thus, we must keep T finite in 
the process of calculation and take the T ^ oo limit after doing the Feynman integrals. 
It turns out that this limit is free of IR divergent terms for the choice of magnetic field 
Eq. (|3.2|) if the path integral is properly regularized. Whether this limit is well-defined 



for more general field strength, than Eq. (3^), is an interesting question we will address 
elsewhere. 

The result is an asymptotic expansion in /i ~ 1/A;, whose coefficients at each order can 
be computed using perturbation theory and depend on local invariants of the metric, such 
as the Riemann and Ricci tensors, curvature scalar and their derivatives. 

Let us begin. To keep track of the T dependence we rescale the time parameter, 
defining 

t = tf + {tf-ti)T = tf+TT 

with T G [—1, 0]. The classical solution for the trajectory with boundary conditions Xi = Xf 
is just a constant function. Introduce normal coordinates z", in the vicinity of the 
classical trajectory 

x" = x'} + z''{t) 



X 



~- = x} + z~^{t) 



Although the gauge, which triviaUzes anti-holomorphic derivatives is rather Aa = (2.7), the difference 
between gauge choices is inessential, since the density matrix is a gauge invariant object. We find it 
convenient to work in the anti-holomorphic gauge. 



-10- 



The normalization factor N can be fixed by considering the standard normahzation of the 



heat kernel in the case of non-coincident initial and final points, as e.g. in |3£], and is equal 
to 

N = k", 

where n is the complex dimension of the manifold. 
3.2 Weyl-ordering counterterm 

The path integral representation (|3.1|) of the heat kernel has been studied since the pio- 



neering work by DeWitt |38|. In the hamiltonian framework the path integral corresponds 
to transition amplitude 

K{xi,xf,T) = {xf\e-^^/''^^\x,). 

Since the kinetic term in H depends on the coordinate variable through the metric, the well 
known subtlety arises in this case with the operator ordering of momentum and coordinate 
variables - different choices of ordering lead to different lagrangians. This issue was studied 



in great detail in [40, 41, ^. There it was shown that there is a convenient choice of the 



ordering in the hamiltonian which preserves general coordinate invariance 

H = \g-^/\p, - A,)G'^G^'\pj - Aj)G-^l^ 

= Ig-'^'Pag'^'gih - A)9''^' + \g'^'\h - h)g-^ gp,g-^l\ (3.3) 

where we specified our hamiltonian to the Kahler case. Here G = deigij = g^ = {deigaa)^- 
To transform the hamiltonian framework to lagrangian we rewrite this expression in a 
Weyl-ordered form (see Appendix B) and then perform the Legendre transform with the 
generalized momenta 

Pa = gab^ ' Pb = S'ab^'^ + ^b- 

The following action, written in euclidean time, appears then in the exponent of path 
integral 

ft 



S = I' dt (^g^^i-2' + A-,^'-^B)j 



The Weyl ordering corresponds to a "mid-point rule" prescription for path integral repre- 
sentation, which will be introduced in the next subsection. The last term in this "quantum 
corrected" action is necessary e.g. to obtain correct path integral representation for the 
small-T heat kernel . In the next section we will see that it is also necessary for obtaining 
the correct infinite-T expansion. 

3.3 Normal coordinates, free action and propagators 

In Kahler normal coordinate frame (see Appendix A for conventions) all pure (anti-) holo- 
morphic derivatives of the metric at a chosen point are set to zero. Setting x = Xi = Xf, 
we use Kahler normal coordinates, centered at x and obtain the following expansions for 
the Kahler potential, metric and gauge connection, up to the sixth order in derivatives 

Kix'^ + z^ + z^) = g,i{x)z'^z' + \K^,-,i{x)z'^ z'z^z' + ^K^,^-^j„{x)z'^ z' z^z^zh' + . . . , 



- 11 - 



1 



in self-explanatory notations. Note that we omitted terms which turn out not to be relevant 
up to the second order in h. For example, the term with five derivatives of mixed type 
K^jj^^l{x)z"'z^z'^z°'z^ is non-zero in our coordinate frame, but it contributes to the density 
matrix only starting from h^, as one can check by power counting. 

Using auxiliary ghost fields 6° and to raise the determinant from the measure ( |3.1] ) 
to the exponent, we can rewrite the diagonal of the density matrix ( p.lj ) as 



p{x) = M 



z(0)=0 



Vz''{t)Vz^{t)VW{t)Vc^{t) e-h 



Iz{-1)=0 

where we split the action into free part 



Sn 



dT 



1 



9ab{x)z"'z^ + kg^i{x)z''z'' + g^bixWc' 



L.b 



^0 ^ Sint 



(3.4) 



and interaction part, which up to the sixth order in derivatives of the Kahler potential, 
looks like 



Si', 



int 



+ k(^K,,,-,{x)z'^z>'z-^ + -^K^ 



1 



+ ( K,,-,i{x)z''z'' + ^K,,^,-,,{x)z'z'z^z' ) b'^c' 



T {R{x) + dcdcR{x)z^z^) 



(3.5) 



and dots denote r derivatives. The propagator for free theory (3.4) 

(#(T)z"(a)) = %'^^A(r,cT), 
satisfies equation 

-Td7^ + M A(r,a)=5(r-.), 
and the path integral boundary conditions translate into boundary conditions for A 
A(-l, a) = A(0, a) = A(t, -1) = A(r, 0) = 

The unique solution is 

A(r.a) = ^-^^ {eir - a)e^^ {I - e^^^){l - e-^^('^+i)) 

+e{a - t)(1 - e-'=^")(l - e^^("+^))} , (3.6) 



- 12 - 



where the step-function is defined using the "mid-point rule" 



( 



1, T > C7 

\, T = a (3.7) 
0, T<a 



This value at zero follows from the choice of symmetric ordering in path integral and is 
crucial for obtaining correct results for heat kernel expansion [p. Ghost propagator can be 
regulated with the help of A(t, cj) in the following way 

{h\a)c\T)) = -hg'^h{T -a) = hg'^' ( ^"A{t, a) - k'A{T, a)) , 



T 



where *A(t, cr) = dA{T,a)/dT, etc. 



3.4 Perturbation theory. First Order 

Now we are ready to study the perturbation theory in h for the diagonal of the density 



matrix (3.4) 



p{x) = Mil + hpi{x) + h^p2{x) + ...). 

From ( p.5| ) the dimension of variable z is h)-/"^ , therefore by power counting pn{x) should 
contain terms with 2n covariant derivatives of the metric. For example, at first order in h 
the only metric invariant is Ricci scalar. 
At the first order in h we have 

hp, = -\k^,-,i jdr + +h^R 

= hR- dT (ACA* + "A) + 'AA') 1^ + h-R = hRIi{T, k) + —R. (3.8) 

and from here on integration always runs from —1 to 0. Here we apply usual Wick rule to 
calculate the correlators and in then we use the fact that Rg^^^i{x) = K^^^i{x) in normal 
frame centered at x. This calculation elucidates the role of the ghost fields 6, c. Namely, 
their contribution cancels the 5(0) terms, which appear in second derivatives of the bosonic 
propagators at coinciding points. 

The values integrals used in the main text and their large time asymptotics are collected 
in Appendix C. In T ^ oo limit of the expression above becomes 



Note, that the Weyl-ordering counterterm ( |B.6| ) is necessary to cancel the large-T diver- 
gence. This calculation provides an independent check of the coefficient in front of this 
term^ (see 0] for detailed consideration of this question) . 



Factor 1/4 here compared to 1/8 in Pf is due to our definition of scalar curvature (A.l) in Kaliler case 



- 13 - 



3.5 Perturbation theory. Second Order 

At the h"^ order the following metric invariants may appear in the expansion: AR = 
g^'^dadaR, |Ricp = i?aa-R"", |Riem|2 = R^-f^f^R''^''^ and R"^. Therefore the second order 
correction splits into four components, corresponding to the listed invariants. The full 
second-order contribution reads 

h'p2 = -l^tc-ai-c Jdr + ^{z^z''z'^-z-^-z'2% + l{z'z^-z^-z%^c%^ 



1 /■/"../ 1 



+^K,-ahK'b'-a'l' j Jdrda ^—{z^z^i-^Xz^'z^'z'^'^'D 

+ ^{zh^z'^i'\yz''z^'i'l) + ^{z'^zh^i\rZ'^'z'''z^'i''\^) 
+ l.{z'^fa-a^b^y-a'j^a'^b'^^^ ^ k{z'^ z'' zH^z'' Z~^' b''' / \„) 
+ {zh%''cXz^' z""' b"' C^' \^)^ 

+^dAR jdr {z'z')\r + ^R-hRh{T,k) + i (^i?)' (3.10) 
We start computation from the first line in this expression. Taking into account the identity 



|A.2|), the first line reads 

-n2(-Ai? + 2|Ric|2 + |Riem|2) jdr ^('AA'A + A^fA' + ••A)/2)|^ 

= -fi^i-AR + 2|Ric|2 + iRiemp) l2{T, k) 

« -;i2(-Ai? + 2|Ricp + |Riem|2) (4^-^ ) , as T ^ oo (3.11) 



Consider now the integral in the second to fifth lines in ( p.lOD . There are several nonequiv- 
alent ways to contract z variables, leading to different invariants. Contraction of each of 
the primed indices a', b' , a', V with a non-primed index, leads to the |Riemp structure. 
Such terms are given by the following expression 

ylRiempy jdrda (^^{A{a,T)A{T,arA'{a,TrA'{T,a) 
+A{a, r)A'(T, a)' A' [a, T)*A(r, a) + •A(ct, t)A{t, a) A* [a, T)'A'(r, a) 
+'A(a, r) A'(r, a) A' (a, T)'A(r, a)) + 2^{A{a, r) A(r, a)' A' {a, r)*A(r, a) 
+'A{a, r) A(r, a) A' (a, t)*A(t, a)) + k^A{(T, r) A(r, CT)'A(a, T)'A(r, a) 
-A{a, r) A(r, a) (^••A(a, r) - k'Aia, r)) (^••A(t, a) - k'A{T, a) 

= y |Riem|2 . /4(T, k) ^ y |Riem|2 - , as T ^ oo (3.12) 

If we contract only two prime and two nonprime indices between each other we get the 
structure |Ricp 

ylRicp f jdrda (^{'A{T)'A{a)'A{T,a)A'{a,T) + 'A{TyA'{a)A{T,a)A-{a,T) 



-14- 



+'A(T)A((j)A'(r,a)'A'(a,T) +'A(r)A'(f7)A(T, a)'A'(a,r) 
+A'(r)'A(a)'A'(r, a)A{a, r) + A'(T)'A'((7)'A(r, a)A{a, r) 
+A*(r)A(CT)*A'(r, a)'A{a, r) + A'(r)A'(a)'A(T, cj)'A(a, r) 
+'A'(r)'A(a)A'(r,a)A(CT,r) +'A'(r)'A'(f7)A(T, cj)A(a,T) 
+'A'(r)A(a)A'(r, a)*A(a, r) + •A'(T)A'(a)A(r, a)'A(a, r) 
+A(T)'A(a)'A'(T,a)A'(a,T) + A(r)'A'(CT)'A(r,a)A'(a,r) 
+A(r)A(a)'A'(r,a)'A'(a,T) + A(r)A'((7)'A(r, a)'A'(c7, r)) 

+-CA(r)'A(a) A(t, a) A' {a, r) + 'Air) A(a) A(r, a)'A'(a, r) 
+A'(r)'A(a)'A(T, cj)A(a, r) + A'(T)A(a)'A(r, a)'A((7, r) 
+'A'(r)'A(a) A(r, a)A{a, r) + •A'(r) A(c7) A(t, (7)'A(a, r) 
+A(T)'A((7)'A(r, a) A* (a, r) + A(r)A(a)'A(r, a)'A'(a, r)) 

+A;2CA(r)'A((7)A(T, a)A{a, r) + •A(r)A(CT)A(T, a)'A(a, r) 
+A(T)'A(cT)'A(T,a)A(a,T) + A(r)A(c7)'A(T, a)'A(a, r)) 

+|CA*(r)A(r,a)A(a,T)+'A(r)A(T,(T)A'(a,r) 
+A'(r)'A(T,a)A(a,T) + A(r)'A(T,a)A'(a,T)) (^"Ala) - rA(a)) 

+2fcCA(r)A(r,(j)A(CT,r) + A(T)'A(r,cj)A(CT,r)) (^^"A{a) - k'A{a) 

+A{T,a)A{a,T) (^"AM - rA(r)) (^"Ala) - rA(a)) 

-A(a)A(r) (^^"A{a,r) - rA(a,r)) (^••A(r,a) - rA(r,a))) 



^|Ric|2 . 15(r, A:) ^ ^iRicp f-^ + A ) , as T ^ oo (3.13) 



If we contract prime indices as well as nonprime indices only between each other, or in 
other words we contract separately z's and 2's at point r and z's and 2's at a, we get 
only disconnected diagrams. The structure of this term is just {hRIif'. Adding up this 
term and last two terms from (3.10) we obtain the first order term ( |3.S| ) squared with the 
coefficient one-half 

-{hpif = - hiRh{T,k) + —R\ « ^R", as T ^ 00. (3.14) 

This term appears since we compute partition function, not the free energy, and therefore 
do not subtract disconnected diagrams. 

Finally the first term in the last line in ( p.lO| ) reads 

^ARjdTA{T,T) = ^ARh{T,k)^h'AR(^-^ + ^y as T ^ 00. (3.15) 



- 15 - 



Let us now collect all the terms ( ^j.ll , 3.12 , 3.13 , 3.15 ) that contribute to p2 and 
compute its T ^ oo limit 

P2 = {hiT, k) + TIsiT, k)/A)AR + (-2/2(r, k) + h{T, k)/2)\mc\^ 
+{-h{T, k) + h{T, A:)/2)|Riem|2 + ^{h{T, k) + T/AfR^ 

« {\aR + T^lRiemp - ^|Ric|2 + \rA , as T ^ oo (3.16) 
fe^ V 3 24 b o / 



Now we are ready write down the full expansion of the density matrix (|3.4D up to second 
order in h 

P = ^\^^^^^% (^^^ + ^iRieml^ - i|Ric|2 + \r'^ + 0{{h/kf )^ . (3.17) 

Note that this expansion is in perfect agreement with the expansion of Bergman kernel, 
obtained in [11|. 

4:. M = (1, 1) Supersymmetry 

4.1 Action, symmetries and propagators 

One can obtain expansions similar to ( ^.17 ) in other quantum mechanical theories. Here we 



consider (1, l)-supersymmetric particle on Kahler manifold with the magnetic field turned 
on. The action is 

S = dt (g^iz'^z' + r{9a-ar + X^dbQa-aV) + + Faa^r) (4.1) 

This action is invariant under the following (1, 1) supersymmetry transformations 

= -eV'" 

= i^e 

= (4.2) 
if the metric is Kahler and if Aa, Aa is a connection of holomorphic vector bundle 

Fab = F-ai = 0. 

Set the field strength to be proportional to the metric, exactly as before in Eq. (p.5|). 
Consider now the path integral representation of this theory. If the boundary conditions 
for X and -0 fields are the same, no ghosts are needed in the action, because bosonic and 
fermionic determinants cancel in the measure. Moreover, no Weyl-ordering counterterm 
is needed in this theory, due to fermions. Bosonic propagator is the same a before, and 
fermionic propagator 



-16- 



satisfies 



da 



+ Tk 



r(T, a) = —5{t — a) 



We would like to compute the "index density" , i.e. the supertrace of the density matrix, 
without performing the x-integral 

p{x) = lim Str(-l)-^e-^-f^. 

T — *oo 

The right hand side here depends only on the bosonic "zero-mode" x, and all fermionic 
dependence is integrated out. Fermion number insertion (—1)^ corresponds to periodic 
boundary conditions for fermions, in which case the propagator has the form 



r(T,a) 



1 



4.2 Perturbation theory 



The idea of the calculation is the same as in nonsupersymmetric case. We use Kahler 
normal coordinates and expand the metric around the constant configuration x. 
Free part of the action is given by 



i: 



dr 



1 



b„la 



(4.3) 



and the interaction part, up to the sixth order in derivatives of the Kahler potential, reads 



Si', 



int 



1 



+ ( K^,-al{x)z''-Z^ + -^K^,,^J„{X)Z'Z'-ZH^ ] ij\Tk + dr)r 



+ ( K,-aiix)z'' + ^-K^,^^J,,{X)Z'Z^Z' ) i'i^'r 



(4.4) 



At the first order in h we get 

p^{M = l) = R jdT (^^('AA* + 'A' A) + fc'AA - A5(0) + AT^ |^ = 0, 

so h} term is exactly zero, even for finite T. 

Computation at the second order in h proceeds in a similar fashion as in previous 
section. Let us only mention one shortcut. Note, that each contraction of ^(o") and 
{Tk + dr)ip{T) is proportional to delta-function 5(t, ex), exactly as contraction of ghosts b 
and c. Therefore the first three lines in the interaction lagrangian (|4.4D generate the same 
terms as bosonic interaction lagrangian ( |3.5| ) and only the last line in ( [4.4D is a new one. 



-17- 



With this observation the calculation simplifies significantly. We only give the final answer 
here 



P2{M = 1) = -{h{T, k) + h{T, k)){-AR + 2|Ricp + |Riem|2) 
+ih{T, k)/2 + /7(r, k) + 78/2)|Ricp 

+(J4(T, k)/2 + h{T, k) - h{T, A;))|Riem|2 (4.5) 

and refer to Appendix C for the values of the integrals here. The coefficients in front of 
|Ricp and |Riemp turn out to be T-independent, as a consequence of supersymmetry and 
the index theorem, and the answer for the density matrix up to the second order in h is 

p{x){N = 1) = fe" (^1 + ^ (2Ai? - |Ric|2 + |Riem|2) + OiK^)^ 

This is consistent with the index theorem. According to the latter the x-integral of 
p{x){M = 1) is equal to the index of Dirac operator on the Kahler manifold M for which 
the exact answer is 

/ dxp{x){N = 1) = indL»A = / chF A i(M). 

Jm Jm 

If we plug F = kg^idz"" A dz^ and expand the A-roof genus A in powers of curvature tensors 
then the first two terms in this expression coincide with first two terms in J p{x){M = 1), 
up to maybe an overall constant. 

5. N — (2, 2) Supersymmetry 

5.1 Action, symmetries and propagators 

The action is 

S = J'' dt (g^iz'^i + i^\{gaar+ + i'diga-ar+) 

+r-{9a-ar-. + x'dhga-ar-) + + Fa-a{r+r+ + V^^V'-)) • (5.1) 

The M = (2, 2) supersymmetry transformations 

fe" = -e+V'+ - e-^^ 
5x^ = -€+^l - e^V'- 

Si^l = 3:% + e_rp^V5^ (5.2) 



5^ =x»6_+e-+r^,V'^V' 

be 



5^^ = ^^e- + e+rf-V'iV^i 



leave the action invariant if connection A is holomorphic and the hermitian Yang-Mills 
equation is obeyed 

g'^'DaF.i = 0. 



- 18 - 



Recall again, that our choice of field strength F^i = kg^^ (|2.5D satisfies this equation. 

The object that we would like to compute is the Dolbeault index density, which corre- 
sponds to taking the supertrace over one species of fermions, and setting the zero modes of 
the second species of fermions to zero. To achieve this, we choose the following propagators 
for the fermions 

{4t{T)r-{<y)) = %"'r-(T,a), 

where r_|_ satisfies periodic boundary conditions: r+( — l, cr) = r+(0, cr), r+(r, — 1) = 
r+(T, 0), and r_ satisfies Dirichlet b.c. r_(—l, cr) = r_(T, 0) = 0. These propagators 
are given by 

r+(r, a) = ^-L^ (e^^(-'^)0(r - a) + e^^^^'^^') e{a - r) 

One also has to add a pair of bosonic ghost fields a"", a"", coming from the path integral 
measure 



5.2 Perturbation theory 



The calculation proceeds along the same lines as in the previous two sections. Here we 
present the final answer for the index density 

p{x) = k''{l + hIn{T, k)R + h^{hi{T, k)AR + (^(r, k)/2 + /n(T, k) + I^iT, k))\mc\'^ 
+ i-hiT, k) + h{T, k)/2 + /io(r, A;))|Riem|2 + ^(r, k)R^ /2) + OiK")) 

«^ A:" (l + + ^ Qa/? + ^|Riem|2 - i|Ric|2 + ^-R^^ + 0{{h/kf)^ , as T ^ oo. 

Notice, that the only T-dependent term here is a total derivative, as is expected from 
the index theorem 

/ dxp{x){M = 2) = mddA= f chFATd(M). 

J M JM 

Recall, that this index formula computes dmi'^^{—\YH^''^{M,L^), which is equal to 
d\m.H^{M,L^) for large enough k. This explains why M = 2 and nonsupersymmetric 
Bergman kernel's expansions coincide. 

6. Conclusions 

In this paper we derived the Tian-Yau-Zelditch et al expansion of the Bergman kernel 
from quantum mechanical path integral. Our results are in complete agreement with the 
calculation of Ref. [11|, using Tian's peak section method. 



-19- 



In quantum mechanics, the Bergman kernel corresponds to density matrix of a particle 
in strong magnetic field on Kahler manifold, projected to the lowest Landau level. The 
expansion in the inverse magnetic flux number can be extracted by taking the infinite time 
T ^ oo limit of the (non-supersymmetric) path integral and using the normal coordinate 
expansion of the metric and the magnetic field. In this paper we considered a the configura- 
tion of magnetic field, being proportional to the Kahler form. Expansions of the same type 
can also be obtained using super symmetric quantum mechanics, where they correspond to 
index densities. It would be interesting to extend this result to the case of particle coupled 



to non-abelian external fields, considered in mathematical literature |31]. 

The physical argument, presented in sec. 2.1, suggests that the analogous expansion for 
the Bergman kernel may be obtained for a more general magnetic field strength, associated 
with holomorphic line bundle. We plan to check this in the future publication. 

One of the most interesting consequences of this result, as discussed in section 2, is 
that there exists a specific magnetic field and metric for which the density matrix 
is constant everywhere on the manifold. This is the "maximally entropic" metric for a 
quantum mechanical observer, in a sense discussed in [p^ j. 

Acknowledgments We would like to thank Z. Lu, S. Lukic, S. Lukyanov, B. Shiffman, 
G. Torroba, K. van den Broek, P. van Nieuwenhuizen and S. Zelditch for useful discussions. 
This work was supported in part by DOE grant DE-FG02-96ER40959. The work of S.K. 
was also supported by the grant for support of scientific schools NSh-3035.2008.2 and RFBR 
grant 07-02-00878. 

A. Curvatures 



We follow conventions of [11 1 



Kabb = dbdlQa-a - g^^dbgacdbdc 



P - — —n^^ R 
-n-aa — tJ ^aabbi 



R = g^'^Ra-a, (A.l) 
AR = g'^'^dad-aR, 
|Riemp = fi^^^^i?""^^, 
|Ric|^ = Raa.R'^'^- 
Let K be the Kahler potential for the metric 

gaa = dadaK. 

In Kahler normal coordinate frame the Cristoffel symbols and all pure holomorphic deriva- 
tives of the metric vanish at the origin 



Kaab^...bAx) = 0, 



-20- 



for any positive integers m, n. The following terms in the Taylor expansions of the Kahler 
potential, the metric, Rieman tensor and Ricci scalar are relevant for the present paper 

Ri^^a ^ ^a^-, ^ ^ ^ i2g'^\x)g'''ix)g'\x)K^,,-,ix)K^,,^ix) 
-g'^-%x)g'\x)K^,^-^j,,{x) - 5'^^(x)5^^(x)/'^"(x)i^,,,,-(x)i^,,,5(x))z^z-" + . . . 

= R{x) + dcdcR{x)z''z^. 
The following useful identity holds in the normal coordinate frame 

9''~^9''Kbc,b-c = -dcB-Mx) + 2R^^R^ + R,i,^R'''%. (A.2) 
B. Hamiltonian 

Here we rewrite the hamiltonian (^]^) in a Weyl-symmetric way. First we simplify the 
expression without the gauge potential 

H = \g-^'Ha9'''9h9-^'^ + \9-^l''h9''^ 9Va9^^l'' = \{Pa9''%+ k9''Ha) (B.l) 

+ ^d-,{g'^^da\ng) + ^da{g''^d-,\ng) + ^g'^^d-.lngdalng, (B.2) 

where we use pa = —ihda, pg, = —ihda- The Weyl ordered form of the first term in the 
previous expression is 

ab-\ ^ i ^ - ah i ^ ah- ,— ah ^ , ab ^ - \ 

[Pa9 Ph)W = ^{PaPb9 +Pa9 Pb + Pb9 Pa + 9 PaPb) 

Therefore 

\iPag'''h+h9''%) = {Pag''%)w + lilPa, [g^'M + \Ph [g^'^Pa]]) (B.3) 

= {Pa9''%)w + ^R + ^9^'ribn,. (B.4) 

The last three terms in ( p. ID can be written as 

B-.ig'^'da Ing) = da{g^'d-, In 5) =-R- 9""'^^.^. (B.5) 
g'''d-,lngdalng = g'^'Tl,rl-. 

Using ( |B.3| , |B.5| ) we get the expression for Weyl ordered hamiltonian (|B.l] ) 

H = {pa9''%)w - ^R. (B.6) 

Now it is straightforward to see that the similar expression holds in the presence of gauge 
connection ( |3.3|) , one just has to shift Pi ^ Pi — Ai in the previous equation. 



- 21 - 



C. Integrals 

Here we collect exact expressions for the integrals that appear in the main text. The 
following short hand notations are used 

•A(t,cj) = dA{T,a)/dT, A'(r,cj) = dA{T,a)/da, "A{T,a) = d^A{T,a)/dT, 
A(r,r) = A(r), 'A{t) = •A(r, C7)|,=,, r(r) = r(r,r) 



and so on. 



h{T, k)= Jdr^ (ACA' + "A) + 'AA*) |, 

= - 4fc(e^T -i)2 {^ + kT + e^^(-2 + kT)) (C.l) 



I,iT,k) = |dr icAA-A + A^CA- + "A)/2)|. = -^^^,^-^-^ (C.2) 
(lO + 3kT + 3e^^(6 + 7kT) + 3e2^^(-6 + 7kT) + e^''^{-W + 3kT)^ 

h{T,k)= JdT A{t, t) = _^^ (2 + kT + e'^{-2 + kT)) (C.3) 

h{T,k)= JJdrda (^■^{A{a,T)A{T,arA'{a,TrA'{T,a) 

+A(a,r)A'(T,a)'A'(a,T)'A(r,a) + •A(a, t)A(t, a)A'(a, T)*A'(r, a) 
+«A(a, t)A«(t, a)A«(a, r)'A(r, a)) + 2|;(A(a, t)A(t, a)'A'(a, r)'A(T, a) 
+*A((7, t)A(t, a) A' (a, r)'A(r, a)) + k^A{a, r) A(t, (t)'A((7, r)'A(T, a) 

-A{a, r) A(r, a) (^"Ala, r) - rA(a, r)) (^"Alr, a) - rA(r, a)) ) 

= ^p^gfcT _ 1)4 + + ISfcTe'^^ + 262^=^ (-7 + 2k'^T^) 

-UkTe^'"'^ + e^''^i7 -2kT)) (C.4) 



h{T,k) = J jdrda (^^(•A(r)«A(a)«A(r,a)A'(a,r) + *A(r)*A'(a)A(r,a)A«(a,r) 

+'A(T)A(a) A'(r, a)*A*{a, r) + •A(r) A*((7) A(t, a)*A*(cj, r) 
+A'(r)'A(a)'A*(r, cj)A(a, r) + A*(r)*A*(cj)*A(r, a)A(a, r) 
+A'(r)A((7)'A'(r, a)'A(a, r) + A*(T)A'((7)*A(r, a)*A(a, r) 
+'A' (r ) 'A (a) A' (r, a) A (fj, r ) + 'A* (r) 'A* (fj) A (r, a) A (a, r) 
+'A'(r)A(c7)A'(r, c7)'A(c7, r) + •A'(T)A'(c7)A(r, a)*A(a,T) 



- 22 - 



+A(T)'A(a)*A'(T, a) A* (a, r) + A(r)'A*(c7)'A(r, a) A' (a, r) 
+A(T)A(a)'A'(r,c7)'A'(c7,T) + A(r)A'(a)*A(T,c7)'A'(a,T)) 

Oh 

+-CA(r)'A(a) A(t, a)A*(a, r) + •A(r)A(a)A(T, a)«A'(a, r) 
+A'(r)*A((T)'A(r, (T)A(c7,r) + A«(r)A((7)*A(T, a)*A((7, r) 
+'A'(r)*A(c7)A(T, a)A{a, r) + •A*(r)A(a)A(r, (7)*A((7, r) 
+A(r)'A(a)*A(r,(7)A'((7,T) + A(T)A((T)'A(r,(7)'A*(a,r)) 
+A;2CA(t)'A(c7)A(t, a) A((7, r) + •A(t) A(a)A(r, (7)*A((7, r) 
+A(r)'A(a)*A(r, (7)A(a, r) + A(r)A(a)*A(r, ayA{a, r)) 

+|;CA«(r)A(r, a)A(a, r) + •A(r)A(r, c7)A'(a, r) 
+A*(T)'A(T,a)A((7,T) + A(r)'A(T,a)A'((7,T) f;^"A(a) - rA(a; 



+2feCA(r)A(r,c7)A(c7,r) + A(r)'A(r,a)A(a,r) (^^"Aia) - k'Aia] 

+A(r,a)A(a,r) (^"AW - fc'A(r)) (^^"A{a) - fc'A(a) 

-A(a)A(r) (^••A(a,r) - rA(a,r)) (^••A(r,c7) - rA(T,a) 

f^2^^kT_i^4 (3 + fcT + 6^=^(4 + 8fcT + k^T^) + 2e^^'^{-7 + k^T'^)+ 

+6^'='^ (4 - S/cT + k^T"^) + e^'='^(3 - /cT)) (C.5) 



/I + p'^-' 
dr TAA'I^ = ^p^_^^gfcr)3 (3 + A;T + 4A;re*^^ + e2*^^(-3 + fcT)) (C.6) 

h{T,k) = j jdrda (r(a,a)(CA*(T)A*(T,(T)A(a,T) + A'(T)'A*(T,c7)A(a,T) 

+*A(r)A*(T, a)A'((7, r) + A(t)'A'(t, a)A'{a, t))/T 
+k{*A{T)A*{T, u)A{a, t) + A(r)*A*(T, a)A{a, r)) 
-5(0)A'(t, a)A{a, r)) + (5(t, a)A'(a) A(T)r(r, a)) 

1 + p'^'^ 

+e2*^^(5 + B/fcT - 2fe2r2) + e2'='^(-5 + 2fer)) (C.7) 

h{T,k) = J Jdrda (A'(r,a)A'(a,r)r(T)r(c7) - A'(r)A'(c7)r(r,a)r(a,T)) 

= 4fc2(_i + ,.T)3 (l + -'^(5 + 4fcT) + e^^^(-5 + AkT) - e^^^) (C.8) 



-23- 



h{T,k) = j Jdrda A*(r,(7)A'(a,r)r(T,a)r(a,r) 



kT 

^ (_(_! + e'=^)2 + A;2TV^) (C.9) 



'10 



{T,k)= j jdrda (^-^A'(t,s)A'(s,t)(r+(t,s)r+(s,t)+r_(t,s)r_(s,t)) 



ji2 

+ — r+(t,s)r+(s,t)r_(i,s)r_(s,t) 



^ ((_l + e^^)2-e*=^fc2T2) (C.IO) 



2(-l + e*^^)4p 



/ii(T,A;) = j jdrda (A*(t, s)A*(s, t)(r+(t) + r_(t))(r+(s) + r_(s)) 

- A'(t)A*(.s)(r+(t,.s)r+(s,t) + r_(t,.s)r_(s,t)) 

- T2(r+(t)r+(s)r_(t,5)r_(s,t) + r_(t)r_(s)r+(t,s)r+(s,t)) 

+ |(r+(.) + r_(.))(A-(t,.)A-(.s,t)-A(t) + (•A-(t) - T5(0))A-(t,s)A(.,t) 
+ (•A'(t,s) - r5(t - s))A*(s,t)A(t) + (•A*(t,s) - r(5(t - s))A(s, t)A'(t)) 
+ 2fe(r+(s) + r_(s))(CA'(t, s) - T5{t - s))A(s, t)A(t) + A'(i, s)A(s, i)*A(t)) 

+ 2rA'(t)(r+(t,s)r+(s,t)r_(s) + r_(t,5)r_(5,t)r+(s))) 

^ r(3(-l + 6^=^)2(1 + e^'^) + A:r(l + e^^'^ {-^ + kT) 



{-l + e'^^^k^ 

+6^=^(5 + A;T) +e2*^^(-3 + 2A;r))) (C.ll) 
h2{T,k) = jdr (^A{A' + kTA){r_ + r+) - ^(('A - T5{0))A^ + 2AA"A) - k'AA^^ |^ 

+e2'=^(-10 + 3A;r)) (C.12) 

in(T, k) = jdr [^{{'^' - TS{o))A + 'AA') + A;'AA - rr+r_ + A'(r+ + r_)^ 1^ 
- Yk (^-^^^ 

7i4(r, k)= jdr (^(('A* - T6mA^ + 2A'AA') + ^'AA^ - TAr+r_ + AA*(r+ + r_ 



-24- 



References 

B. S. DeWitt, "Dynamical theory of groups and fields," Gordon and Breach, New York, 
(1965). 

P. B. Gilkey, "Invariancc theory, the heat equation and the Atiyah-Singer index theorem," 

Publish or Perish, Wilmington, Delware, (1984). 

If. Kleinert, "Path Integrals in Quantum Mechanics, Statistics, Polymer Physics, and 
Financial Markets," World Scientific, Singapore, (2004). 

F. Bastianelli and P. van Nieuwenhuizen, "Path integrals and anomalies in curved space," 
Cambridge University Press, UK, (2006). 

L. Alvarez-Gaume and E. Witten, Nucl. Phys. B 234, 269 (1984). 
L. Alvarez-Gaume, Commun. Math. Phys. 90, 161 (1983). 
D. Friedan and P. Windey, Nucl. Phys. B 235, 395 (1984). 

G. Tian, "On a set of polarized Kahler metrics on algebraic manifolds," J. Differential Geom. 
32 (No. 1) 99-130 (1990). 

S. Zelditch, "Szego kernels and a theorem of Tian," Internat. Math. Res. Notices (No. 6) 
317-331 (1998). 

D. Catlin, "The Bergman kernel and a theorem of Tian," in "Analysis and geometry in 
several complex variables (Katata, 1997)", Trends Math. p. 1-23, Birkhduser Boston, 
Boston, (1999). 

Z. Lu, "On the lower order terms of the asymptotic expansion of Tian-Yau-Zelditch," Amer. 
J. Math. 122 (No. 2) 235-273 (2000). 

S. K. Donaldson, "Scalar curvature and projective embeddings. I" , J. Differential Geom. 59 
(No. 3) 479-522 (2001). 

S. K. Donaldson, "Scalar curvature and projective embeddings. 11", Q. J. Math. 56 (No. 3) 

345-356 (2005), [arXiv:math.DG/0407534]. 

S. K. Donaldson, "'Some numerical results in complex differential geometry," 
[arXivmath.DG /0512625]. 

X. Ma and G. Marinescu, Holomorphic Morse Inequalities and Bergman Kernels, Progress in 
Mathematics, Vol. 254, Birkhauser, (2007). 

F. A. Berezin, "Quantisation of Kahler manifold," Comm. Math. Phys. 40 153 (1975). 

N. M. J. Woodhouse, "Geometric quantization," Oxford University Press, Oxford, (1992). 

J. H. Rawnsley, "Coherent States and Kahler Manifolds," Quart. J. Math. 28 403-415 (1977). 

N. Reshetikhin and L. Takhtajan, "Deformation quantization of Kahler manifolds," in L.D. 
Faddeevs Seminar on Mathematical Physics, Amer. Math. Soc. Transl. Ser. 2, Vol. 201 
257276 (2000), [math.QA/9907171]. 

J. R. Klauder and E. Onofri, "Landau levels and geometric quantization," Int. J. Mod. Phys. 

A 4 3939 (1989). 

W. D. Kirwin "Coherent states in geometric quantization," J. Geom. Phys. 57 2 531-548 
(2007). 



-25- 



[22] S. Lukic, "Balanced metrics and noncommutative Kahler geometry," arXiv:0710.1304 

[hcp-th] . 

[23] C. luliu-Lazaroiu, D. McNamee and C. Saemann, "Generalized Berezin quantization, 
Bergman metrics and fuzzy Laplacians," arXiv:0804.4555 [hep-thj. 

[24] M. R. Douglas and S. Klevtsov, "Black holes and balanced metrics," to appear. 

[25] S. Girvin, "The Quantum Hall Effect: Novel Excitations and Broken Symmetries," 

Topological Aspects of Low Dimensional Systems, ed. A. Comtet, T. Jolicoeur, S. Ouvry, F. 
David Springer- Verlag, Berlin and Les Editions de Physique, Les Ulis, (2000). 

[26] R. lengo and D. p. Li, Nucl. Phys. B 413, 735 (1994) [arXiv:hep-th/9307011]. 

[27] S. C. Zhang and J. p. Hu, "A Four Dimensional Generalization of the Quantum Hall Effect," 
Science 294, 823 (2001) [arXiv:cond-mat/0110572]; "Collective excitations at the boundary 
of a 4D quantum Hall droplet," [arXiv:cond-mat/0112432]. 

[28] H. Elvang and J. Polchinski, [arXiv:hep-th/0209104]. 

[29] D. Karabali and V. P. Nair, Nucl. Phys. B 641, 533 (2002) [arXiv:hep-th/0203264]; Nucl. 
Phys. B 679, 427 (2004) [arXiv:hep-th/0307281]; Nucl. Phys. B 697, 513 (2004) 
[arXiv:hep-th/0403111]. 

[30] D. Karabali and V. P. Nair, J. Phys. A 39, 12735 (2006) [arXiv:hep-th/0606161]. 

[31] X. Dai, K. Liu and X. Ma, "On the asymptotic expansion of Bergman kernel," C. R. Acad. 
Sci. Paris, Ser. I, 339 (2004) [arXiv:math/0404494]; X. Ma and G. Marinescu, "Generalized 
Bergman kernels on symplcctic manifolds," Adv. in Math. 27 L 4 1756-1815 (2008); 
[arXiv:math/0411559]; X. Ma and G. Marinescu, "Holomorphic Morse Inequalities and 
Bergman Kernels," Progress in Mathematics, Birkhduser Vol. 254 (2006). 

[32] R. Berman, B. Berndtsson and J. Sjostrand, "A direct approach to Bergman kernel 
asymptotics for positive line bundles," [arXiv:math/0506367]. 

[33] E. Witten, "Supersymmetry and Morse theory," J. Diff. Geom. 17 4 661-692 (1982); 

J-P. Dcmailly, "Champs magnctiqucs ct inegalite de Morse pour la d-cohomologic," Ann. 
Inst. Fourier 355 185 229 (1985); J-M. Bismut, "Demailly's asymptotic Morse inequalities: a 
heat equation proof," J. Funct. Anal. 72 2 599-620 (1987); R. Berman, "Bergman kernels 
and local holomorphic Morse inequalities," [arXiv:math/0211235]. 

[34] H. Luo, "Geometric criterion for Mumford-Gieseker stability of polarized manifold," J. 
Differential Geom. 49 (No. 1) 577-599 (1998). 

[35] J. P. Bourguignon, P. Li and S. T. Yau, "Upper bound for the first eigenvalue of. algebraic 
submanifolds," Comment. Math. Helvetici 69 199-207 (1994). 

[36] M. R. Douglas, R. L. Karp, S. Lukic and R. Reinbacher, J. Math. Phys. 49, 032302 (2008) 
[arXiv:hep-th/0612075]. 

[37] M. R. Douglas, R. L. Karp, S. Lukic and R. Reinbacher, JHEP 0712, 083 (2007) 

[arXiv:hop-th/0606261]. 

[38] B. S. DeWitt, Rev. Mod. Phys. 29, 377 (1957). 

[39] F. Bastianelli and O. Corradini, "On mode rcgularization of the configuration space path 
integral in curved space," Phys. Rev. D 60, 044014 (1999) [arXiv:hep-th/9810119]. 



-26- 



[40] F. BastiancUi, "The Path integral for a particle in curved spaces and Weyl anomalies," Nucl. 
Phys. B 376, 113 (1992) [arXiv:hep-th/9112035]. 

[41] F. Bastianelli and P. van Nieuwenhuizen, "Trace Anomalies From Quantum Mechanics," 
Nucl. Phys. B 389, 53 (1993) [arXiv:hep-th/9208059]. 



-27-